Upper Cervical Spine Loading Simulating a Dynamic Low-Speed …winkelst/Holsgrove2016.pdf · 2016....

Timothy P. HolsgroveDepartment of Bioengineering,

School of Engineering and Applied Science,

University of Pennsylvania,

210 South 33rd Street,

Room 240 Skirkanich Hall,

Philadelphia, PA 19104

e-mail: [email protected]

Nicolas V. JaumardDepartment of Neurosurgery,

Pennsylvania Hospital,


Washington Square West Building,

235 South 8th Street,



Nina ZhuDepartment of Bioengineering,







Nicholas S. StiansenDepartment of Bioengineering,







William C. WelchDepartment of Neurosurgery,







Beth A. Winkelstein1

Department of Bioengineering,

School of Engineering

and Applied Science,




Philadelphia, PA 19104;

Department of Neurosurgery,







Upper Cervical Spine LoadingSimulating a Dynamic Low-Speed Collision SignificantlyIncreases the Risk of PainCompared to Quasi-StaticLoading With Equivalent NeckKinematicsDynamic cervical spine loading can produce facet capsule injury. Despite a large pro-portion of neck pain being attributable to the C2/C3 facet capsule, potential mechanismsare not understood. This study replicated low-speed frontal and rear-end traffic collisionsin occiput-C3 human cadaveric cervical spine specimens and used kinematic and full-field strain analyses to assess injury. Specimens were loaded quasi-statically in flexionand extension before and after dynamic rotation of C3 at 100 deg/s. Global kinematics inthe sagittal plane were tracked at 1 kHz, and C2/C3 facet capsule full-field strains weremeasured. Dynamic loading did not alter the kinematics from those during quasi-static(QS) loading, but maximum principal strain (MPS) and shear strain (SS) were signifi-cantly higher (p¼ 0.028) in dynamic flexion than for the same quasi-static conditions.The full-field strain analysis demonstrated that capsule strain was inhomogeneous, andthat the peak MPS generally occurred in the anterior aspect and along the line of the C2/C3 facet joint. The strain magnitude in dynamic flexion continued to rise after the rota-tion of C3 had stopped, with a peak MPS of 12.52 6 4.59% and a maximum SS of5.34 6 1.60%. The peak MPS in loading representative of rear-end collisions approachedmagnitudes previously shown to induce pain in vivo, whereas strain analysis using linearapproaches across the facet joint was lower and may underestimate injury risk comparedto full-field analysis. The time at which peak MPS occurred suggests that the decelerationfollowing a collision is critical in relation to the production of injurious strains withinthe facet capsule. [DOI: 10.1115/1.4034707]

Keywords: cervical spine, facet, injury, dynamic, strain

1Corresponding author.Manuscript received February 10, 2016; final manuscript received September 8,

2016; published online November 3, 2016. Assoc. Editor: Brian D. Stemper.

Journal of Biomechanical Engineering DECEMBER 2016, Vol. 138 / 121006-1Copyright VC 2016 by ASME

Downloaded From: http://biomechanical.asmedigitalcollection.asme.org/ on 11/06/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

1 Introduction

Dynamic loading of the upper cervical spine can lead to injury,instability, and pain [1–3]. Neck injury is the most common injuryin motor vehicle occupants requiring hospital emergency depart-ment treatment in the U.S. [4]. It has been estimated that betweenone-quarter and one-third of people exposed to rear-end collisions,which commonly occur at speeds of 15–30 km/h, experience neckinjury [5], and that 15–40% of those injuries will result in chronicneck pain, which can extend to the head, shoulders, and arms[6–8]. Although chronic neck pain is prevalent among the generalpopulation [9,10], exposures like those due to low-speed collisionshave been reported to increase the incidence by 2.7 times [11].

The origin of a variety of pain symptoms can be traced to spe-cific regions of the spine [9,12,13]. Nerve blocks and/or injectionsand provocative studies demonstrate that for 35–60% of peoplewith chronic neck pain, the source of the pain is the cervical facetjoint [3,10,12,14]. Moreover, the C2/C3 and C5/C6 cervical spinallevels are those most commonly identified as painful in both idio-pathic neck pain [12] and in chronic neck pain from whiplashexposure [3,14].

Any disruption to the various hard and soft tissue structures ofthe facet joint has the capacity to elicit pain [2,15,16]. The facetcapsule and synovial folds are innervated by nociceptive andmechanoreceptive afferents [17–21]. Pain can result from directdamage of nociceptors [22–24], but can also be produced indi-rectly through damage to the mechanoreceptors, which altersfeedback and increases neck instability, leading to pain in musclesand/or from muscular contractions [1,22,25–27].

Previous studies of facet injury from dynamic neck loading,such as whiplash exposure, have focused on defining the globaland local kinematics of the cervical spine [28–30], deformation ofthe facet capsule [25,31–33], and acute and chronic pain responses[24,34,35]. Human volunteer studies simulating the rear-end colli-sions define a characteristic response, with the torso moving priorto the head, leading the neck to undergo an S-shaped configurationapproximately 60–100 ms after vehicle impact, followed by therearward rotation of the head extending the neck approximately85–140 ms after impact, before the head and torso rebound due todeceleration following the impact, combined with the effect ofseat support and muscle activation [30,36]. Simulations of rear-end collisions using human cadaver models demonstrate similarkinematic patterns over the same time period [28,37–39] and alsoshow that the facet capsule linear strain increases at C3/C4 and

C6/C7 with an impact acceleration of 8 g compared to a physio-logical rate of spine loading [25]. More detailed strain measure-ments of the lower human cervical spine have been estimatedusing stereophotogrammetry to define the three-dimensional (3D)full strain-field of the facet capsule and demonstrate that physio-logical loading of the spine can induce strains sufficient to induceinjury [32,40]. A key finding of both the human volunteer andcadaveric studies is that intervertebral rotations do not exceed thenormal physiological range and that injuries result from abnormalkinematics, such as altered intervertebral rotation axes, which canincrease the loading in and across the facet joint [25,36].

The subcatastrophic and catastrophic failure of human facetcapsules is highly variable [32,40–43]. Despite this, subcata-strophic strains of 35–67% are associated with injury and/or injurypotential [32,40,41], which is within the range measured duringin vitro studies of rear-end collisions in the lower cervical spine[25]. But these studies do not account for the large proportion ofexposures in which pain is attributed to the upper cervical spine,and specifically the C2/C3 level [3,14]. In vivo studies of the neu-ral activity in the facet capsule of goats indicate that the facet cap-sule has both low- and high-strain threshold units, whichdischarge at 10% and 47% strain, respectively [44]. Some low-threshold units are likely to play a role in proprioception, but cap-sular strains as low as 19% have been shown to induce pain in theabsence of any failure of the facet capsule [35] when imposed atrates (500%/s) comparable to those sustained during whiplashexposures [45,46]. As such, defining facet strains in the context ofthe potential for pain and relative to typical mechanical metrics oftissue failure is necessary to understand injury risk in humans.

Despite the prevalence of neck injury resulting from dynamicloading events, and the large proportion of such injuries beingreported in the upper cervical spine [3,14], most investigations ofthe facet capsule response have been focused in the lower cervicalspine. The full strain-field of the C2/C3 facet capsule during anyloading exposure, regardless of rate, has not yet been defined norhave the spinal kinematics been used to contextualize facet behav-ior during potentially injurious exposures with respect to equiva-lent vertebral rotations during normal physiological loading.Accordingly, this study uses a human cadaver model to investi-gate the response of the C2/C3 facet to dynamic flexion–extensionloading of the upper cervical spine and compares the vertebralkinematics and full-field strains at C2/C3 with those responsesunder quasi-static loading.

Fig. 1 Global view (a) of the occiput (Occ)-to-C3 specimen, with tracking markers at eachlevel. The Occ was fixed to a phantom head. The C3 vertebra was rigidly fixed to the cradle,which was actuated for dynamic tests, with all other levels unconstrained in the sagittal andaxial planes. The C2/C3 facet capsule was imaged (b) with two cameras, from which 2D facetkinematics were measured (c), and 3D reconstructions were used to measure the maximumprincipal strain (MPS) (d) and shear strain.

121006-2 / Vol. 138, DECEMBER 2016 Transactions of the ASME


2 Materials and Methods

2.1 Specimen Preparation. Fresh frozen occiput-C3 speci-mens (n¼ 6; three males and three females), 66 6 7 years of age,were used in this study, having been acquired by appropriatemethods and procedures for our institution. Prior to testing, thespecimens were thawed overnight at room temperature while insealed bags; all the musculature was dissected and only the disks,ligaments, and facet capsules were left intact. Self-tapping screwswere driven into the occiput, and self-tapping screws combinedwith Kirschner wires were driven into the vertebral body and spi-nous process of C3 to fix the specimen into pots along with a two-part fast-curing liquid polymer (Smooth Cast

VR

300, Smooth-On,Inc., Macungie, PA). The C3 vertebra was potted with the centerof the vertebral body aligned with the superior end of the speci-men pot, and with an anterior rotation of 25 deg, based on the nat-ural angular orientation of that vertebra in the normal lordosiswith the head in the neutral position [47]. A surrogate head with amass (4.15 kg) and moment of inertia in the sagittal plane(0.0210 kg m2) representative of an average human head [48] wasfixed to the occiput specimen potting fixture to account for theinertial effects of the head mass during loading. The center ofgravity of the surrogate head was located approximately 60 mmabove, and 20 mm anterior to, the occipital condyles [48].

2.2 Test System. Tests were performed using a custom-developed cradle assembly actuated via a servo-hydraulic testingmachine (370.02 FlexTest 60, MTS Systems Corp., Eden Prairie,MN) (Fig. 1). A linkage between the servo-hydraulic actuator andthe edge of the cradle assembly converted linear motion in theactuator into angular motion in the cradle assembly. The center ofrotation of the cradle assembly was located in a fixed positionapproximately 16 mm below the center of the C3 vertebral bodyfor each specimen when in the neutral position; this orientationintroduced a combination of rotation and translation to the C3 ver-tebral during dynamic tests. The actuator of the testing machinewas fitted with a 5 g capacity uniaxial accelerometer (Model7521A2, Dytran Instruments, Inc., Chatsworth, CA) to measurethe acceleration and deceleration during dynamic tests. Fourmarkers were attached to the surrogate head and three additionalmarkers were attached to the loading cradle to track their motionsin order to obtain the rotation of the C3 vertebra, and the globalrange of motion (ROM) of the specimen. Distracting pins (12 mmthreaded head) with four reflective polypropylene markers werescrewed into the C1 and C2 in an anterior–posterior orientation(Fig. 1(a)), which were used for motion tracking. The C2/C3 facetjoint-line was identified via visual inspection and gentle manipula-tion of the joint, and 1 mm opaque-black painted steel beads werefixed in an array over the outer surface of the entire capsule usinga minimal amount of cyanoacrylate glue (Fig. 1(b)). Positioningof the marker array with respect to the joint-line was confirmedusing lateral radiographs of each specimen. The C3 was rigidlyfixed to the loading cradle (Fig. 1). Pins attached to the surrogatehead fitted between two acrylic guide rails, which constrained thespecimen in the coronal plane, but allowed free movement in thesagittal and axial planes.

One high-speed camera (V4, Vision Research, Inc., Wayne,NJ), with 256� 256 pixel resolution, tracked the two-dimensional(2D) global kinematics of the specimen in the sagittal plane,which enabled measuring the vertebral marker displacement andthe intervertebral rotation at each level. The images were cali-brated against the known distance between the markers on theactuation cradle, providing a scale of 0.87 6 0.02 pixels/mm.High-speed stereophotogrammetry was adopted using two cam-eras (Miro eX1; 480� 360 pixel resolution, Vision Research,Inc.), which tracked the marker array on the C2/C3 facet capsuleto enable the 3D reconstruction of facet capsule deformation(Fig. 1(d)). The dual-camera array was calibrated prior to testingeach specimen using a standard marker array and had an error of0.16 6 0.12 mm. All the cameras were synchronized; the position

and acceleration of the actuator on the MTS and all the imagedata were acquired at 1 kHz for all the tests.

2.3 Test Protocol. Each specimen underwent quasi-static anddynamic testing in flexion and extension: quasi-static testing repli-cated normal physiological flexion–extension, and the dynamictests simulated low-speed front- and rear-end collisions. Quasi-static tests were completed before and after dynamic tests toassess if there were any effects of the dynamic loading on verte-bral kinematics. A 5-min recovery period was allowed betweeneach test, with the specimen fixed in the neutral position.

The quasi-static tests were performed by commencing dataacquisition prior to manually guiding the surrogate head throughflexion and extension with C3 remaining stationary. The globalROM applied to specimens was 16–20 deg in flexion and13–20 deg in extension. Dynamic loading consisted of the surro-gate head being held in the neutral position by an electromagnet,which was released simultaneously with the actuation of the C3vertebra via the loading cradle. The loading cradle was used toapply 6.1 deg of rotation to C3 at 100 deg/s and approximately1.7 mm of translation, which replicated the rotation of the C3 ver-tebra during a low-speed collision [49]. The actuation was appliedusing a square wave, with the accelerometer on the actuator usedto determine the acceleration and deceleration to/from the test rateequating to 100 deg/s. A separate dynamic test was performed toactuate C3 in flexion combined with anterior translation, andextension combined with posterior translation, which replicatedrear- and front-end collisions, respectively. A trigger was used tosimultaneously release the electromagnet and actuate the loadingcradle, as well as to initiate data acquisition.

2.4 Data Analysis. The displacements of the markers defin-ing the vertebral kinematics and the facet capsule deformationwere analyzed for all the tests using ProAnalyst software (Version1.5.7.7, Xcitex, Inc., Woburn, MA). The 3D displacement data ofthe markers on the facet capsule were used to estimate the full-field Lagrangian strain using LS-DYNA software (Version R7.0.0,Livermore Software Corporation, Livermore, CA), with the posi-tion of the markers 20 ms prior to actuation serving as the 0%strain reference condition.

The two sets of quasi-static flexion–extension tests for eachspecimen were compared at each vertebral level using 1 deg incre-ments of global ROM from 16 deg of flexion to 13 deg of exten-sion. Comparisons of the rotation at each cervical level (Occ/C1,C1/C2, and C2/C3) relative to the global ROM were made usingtwo-way repeated measures ANOVAs (IBM SPSS Statistics22.0.0.1, IBM Corporation, Armonk, NY) with a significancelevel of 0.05. The 3D full-field strains during dynamic loadingwere assessed over a 150 ms window, with the actuation occurringat 20–81 ms. The maximum principal strain (MPS) and shearstrain (SS) were calculated for each element of the facet capsulearray at 10 ms intervals. In order to include the strain behavior atthe time that the actuation was stopped, measurements at 81 msafter the test started (61 ms after C3 actuation began) were used inpreference to the 80 ms time interval.

All the comparisons of the MPS and SS between dynamic andquasi-static loading were performed using Wilcoxon signed-ranktests (IBM SPSS Statistics) with significance at 0.05. For eachspecimen, the peak MPS during dynamic loading was identifiedand compared with the peak MPS during quasi-static loading atequivalent C2/C3 rotation. In addition, for each specimen, themean MPS across all the elements of the facet capsule at the timeof peak MPS was calculated and compared with the mean MPSduring quasi-static loading to assess whether the peak capsulestrains were similar to the overall facet capsule behavior. In addi-tion to statistical comparisons of MPS magnitude, the directionand location of the peak MPS on the facet capsule were identifiedto assess the strain characteristics between specimens, and loadingconditions. The maximum and minimum SS during dynamic

Journal of Biomechanical Engineering DECEMBER 2016, Vol. 138 / 121006-3


flexion and extension were also compared with the SS duringquasi-static loading at equivalent C2/C3 rotations. To account forthe curvature of the facet capsule, the SS was estimated in all thethree anatomical planes, and the peak value was taken as the max-imum. Both the maximum and minimum SS were measured toaccount for positive and negative shear occurring at different peri-ods within the exposures. Comparisons were made between themaximum and minimum SS during dynamic testing with equiva-lent quasi-static data using Wilcoxon signed-rank tests (IBMSPSS Statistics) and significance at 0.05.

In order to relate the facet behavior under dynamic loading topreviously published in vitro cadaveric studies, the position of themost superior and inferior rows of markers was used to calculatethe 2D facet joint posterior–anterior sliding, compression–separation, and linear strain across the C2/C3 facet using methodspreviously described [25] (Figs. 1(b) and 1(c)). Posterior slidingwas taken as positive, facet separation was defined as the maxi-mum distance across the facet, compression defined as the mini-mum distance across the facet, and the linear facet strain wasdefined as the distance between corresponding pairs of markers onthe superior and inferior facets relative to the distance 20 ms priorto actuation of C3. The maximum posterior–anterior sliding,separation–compression, and linear strain across the facet eachwere calculated at the same time increments as the full-field strainanalysis, and statistical comparisons were made with equivalentquasi-static data using Wilcoxon signed-rank tests (IBM SPSSStatistics) with significance at 0.05.

3 Results

Dynamic flexion–extension did not significantly alter the inter-vertebral rotations compared to quasi-static loading at any cervi-cal level (p¼ 0.997 for Occ/C1; p¼ 0.999 for C1/C2; andp¼ 0.714 for C2/C3) (Fig. 2). The average acceleration/decelera-tion of the MTS actuator to/from the test rate equating to 100 deg/s during dynamic tests was 2.8 6 0.8 g and 2.8 6 0.8 g in dynamicflexion and 1.8 6 0.2 g acceleration and 2.8 6 0.3 g deceleration indynamic extension. All the specimens exhibited similar kinematicpatterns at each level during dynamic loading in flexion andextension, although the magnitude of rotation at each level didvary between specimens (Fig. 3). Dynamic loading altered therotation at the C2/C3 level from extension to flexion in the flexiontests and from flexion to extension in the extension tests (Fig. 3).This change occurred shortly after the actuation stopped at 81 ms.After 150 ms during the dynamic tests, all the specimens movedinto flexion due to the moment resulting from the mass and theposition of the center of gravity of the surrogate head. The maxi-mum and minimum rotation at each level was generally less thanthe rotation during quasi-static flexion–extension tests (Table 1)and was within physiological ROMs previously reported [50,51].

In dynamic flexion, the peak MPS in the capsule was12.52 6 4.59% and the mean MPS was 6.59 6 2.93%. In dynamic

extension, these strains were 7.02 6 1.88% and 2.08 6 0.89%,respectively. The peak MPS occurred at the time when dynamicactuation stopped (81 ms) or later in all the flexion tests, but wasmore varied when extension was applied, occurring both beforeand after actuation was stopped (range 40–150 ms) (Fig. 4). Thepeak MPS was significantly higher during dynamic flexion thanduring quasi-static loading (p¼ 0.028), whereas there was no dif-ference between peak MPS in dynamic extension and quasi-staticloading (p¼ 0.600) (Fig. 5(a)). The same statistical outcomeswere found between the mean MPS during dynamic and quasi-static loading. The mean MPS was significantly higher in dynamicflexion than during quasi-static loading (p¼ 0.028), but there wasno difference between dynamic extension and quasi-static loading(p¼ 0.600) (Fig. 5(b)).

The location of the peak MPS was in the center and toward theanterior aspect of the facet capsule in dynamic flexion (Fig. 6). Indynamic extension, it was largely located in the center and towardthe posterior aspect of the capsule. The direction of the peak MPSwas orientated approximately along the facet joint, particularly inthe region of capsule where the maximum strain was sustained.The general position and direction of the MPS were similar todynamic conditions at equivalent quasi-static rotations, but themagnitude was less pronounced relative to the surroundingelements (Fig. 6).

The greatest shear strain generally occurred in the sagittal planein all the specimens throughout dynamic loading, although thosewith more pronounced curvature of the facet exhibited maximum/minimum SS in the coronal and axial planes at some time incre-ments. In dynamic flexion, the maximum SS was 5.18 6 1.48%and the minimum was �4.30 6 2.73%. In dynamic extension,these strains were 2.65 6 1.65% and �4.38 6 1.50%, respectively.The greatest SS changed from a minimum to a maximum indynamic flexion, with the reverse occurring in dynamic extension;these changes corresponded approximately to the time when theactuation of C3 stopped (at 81 ms) (Fig. 4). One specimen wasomitted from the comparison of the maximum SS in dynamic flex-ion, and a separate specimen was omitted from the dynamic exten-sion comparison due to no C2/C3 rotations during quasi-staticloading being within 0.1 deg of the C2/C3 rotation duringdynamic loading. The maximum SS was significantly greater indynamic flexion (p¼ 0.028) and extension (p¼ 0.028) comparedto quasi-static loading, but there was no difference in the mini-mum SS (Table 2).

The facet sliding was consistent across all the pairs of markersfrom the anterior to posterior aspect of the facet, so the magnitudewas averaged at each time increment for each specimen (posteriorsliding taken as positive). The separation, compression, and linearstrain varied across the facet; therefore, the greatest value at eachtime point was used for analysis. Posterior–anterior facet slidingexhibited similar behavior to the rotational kinematics, with thefacet undergoing posterior sliding during actuation in dynamicflexion and changing to anterior sliding at approximately the time

Fig. 2 Mean (695% confidence intervals (CI)) quasi-static kinematics (flexion positive) at the Occ/C1 (a), C1/C2 (b), and C2/C3(c) levels with respect to the global ROM predynamic (Pre-D) and postdynamic (Post-D) loadings



actuation stopped; the reverse was observed in dynamic extensionloading. Some specimens were excluded from the statistical com-parisons, since no C2/C3 rotations during quasi-static loadingwere within 0.1 deg of the C2/C3 rotations of dynamic loading(Table 3). The maximum posterior–anterior sliding was�1.07 6 0.42 mm in dynamic flexion and 1.13 6 0.43 mm indynamic extension. The separation–compression was less

consistent, but remained within 0.53 mm of separation and0.61 mm of compression relative to the neutral position for all thespecimens in all the tests. The maximum linear strain across thefacet was 3.92 6 1.33% in dynamic flexion and 2.58 6 0.98% indynamic extension. Dynamic loading produced significantly moreposterior–anterior sliding than quasi-static loading (p� 0.043)(Table 3). The maximum linear strain across the facet joint was

Fig. 3 Mean (695% CI) rotation angle (flexion positive) at Occ/C1 (a), C1/C2 (b), and C2/C3 (c) levels during dynamic flexion,and dynamic extension (d)–(f) applied at 100 deg/s to the C3 level from 20 to 81 ms

Fig. 4 Mean (695% CI) peak MPS ((a) and (b)) and peak SS ((c) and (d)) of the C2/C3 facet cap-sule during dynamic actuation of the C3 applied at 100 deg/s from 20 to 8 ms in flexion ((a) and(c)) and extension ((b) and (d))



lower than the peak MPS from the full-field capsular analysis, andno differences were found between the linear facet strain in eitherdynamic flexion or dynamic extension compared to quasi-staticloading (p� 0.345) (Table 3).

4 Discussion

This study compares the cervical spine kinematics, joint kine-matics and strains, and full-field strain of the C2/C3 facet capsuleduring dynamic flexion–extension with corresponding physiologi-cal ROMs. The dynamic loading profiles simulate those sustainedby the neck of a seated occupant during rear-end and frontal colli-sions [49]. Although the vertebral rotations during dynamic load-ing did not exceed physiological ROMs (Table 1) [50,51], boththe MPS and SS in the facet capsule were increased duringdynamic flexion relative to those sustained during correspondingquasi-static rotations (Fig. 5; Table 2). The 100 deg/s rotation rateused in the present study was selected to replicate the rotation ofthe C3 during a low-speed collision [49,52], from which neckinjury can commonly result. Further, the acceleration of loading(1.8–2.8 g) was in line with the previously reported data [25,49].

Muscle activity was not simulated during this study, althoughthis is likely to have minimally affected the results. Muscle activ-ity has been shown to be limited to an activation response time ofapproximately 51 ms and the time to maximum force production

for neck muscles is approximately 114 ms [53], the time periodsover which the C3 vertebra was actuated (61 ms) and over whichdata were analyzed (130 ms) would not incorporate muscle contri-butions. The intervertebral rotations during dynamic flexion werecomparable to those previously reported (Table 4). The relativelylarge rotations, particularly in extension reported by Grauer et al.[54], may be the result of the flexion moment due to the mass ofthe head being was counteracted by a pneumatic suspension sys-tem. The study of Ivancic et al. [39] used muscle force simulationto stabilize the head, which resulted in similar peak flexion andextension rotations to the present study. While a stabilizing pre-load could have been applied to simulate passive muscle activity,it was not used in the present study. Although this is a limitationas the stability that muscle forces would provide in vivo wasabsent in these experimental conditions, the results of this studycould be taken as a worst-case scenario of the neck response tofront- and rear-end collisions. Further research into the effect ofmuscle forces on the cervical spine and facet strain responses dur-ing dynamic loading exposures would help to assess the likelihoodof injury under different loading conditions.

The facet kinematics in this study exhibit similar behavior tothose previously reported, with the actuation of the C3 vertebracausing posterior sliding of the C2 level relative to the actuatedlevel [25,49,55]. However, the greatest sliding during rear-endcollision simulation in our study was anterior sliding (Table 3),

Fig. 5 Mean (695% CI) peak MPS (a) and mean MPS (b) on the C2/C3 facet capsuleduring flexion and extension at 100 deg/s compared to quasi-static loading. An aster-isk (*) denotes a significant difference (p < 0.05).

Table 2 Mean 6 SD SS in the C2/C3 facet capsule during dynamic flexion and extension exposures and equivalent C2/C3 rotationduring quasi-static loading

Test Parameter n 100 deg/s Quasi-static Significance

Flexion Maximum SS (%) 5 5.34 6 1.60 1.73 6 2.20 0.028a

Minimum SS (%) 6 �4.30 6 2.73 �4.69 6 5.55 0.917Extension Maximum SS (%) 5 2.73 6 1.83 2.43 6 2.10 0.028a

Minimum SS (%) 6 �4.38 6 1.50 �3.37 6 5.49 0.463

aA significant difference between dynamic and quasi-static tests (p< 0.05).

Table 1 Mean 6 standard deviation (SD) maximum and minimum rotations (flexion positive) at each cervical spinal level duringdynamic flexion and extension, and quasi-static rotation

Flexion Extension Quasi-static

Level Max Min Max Min Max Min

Occ/C1 5.31 6 2.76 �0.39 6 0.37 0.80 6 1.04 �7.73 6 3.30 5.20 6 2.92 �13.11 6 4.78C1/C2 7.88 6 3.53 �0.02 6 0.03 0.96 6 1.20 �1.81 6 0.54 14.60 6 5.26 �3.27 6 1.52C2/C3 1.74 6 1.14 �1.54 6 0.98 2.04 6 2.08 �2.17 6 0.65 3.71 6 4.40 �4.06 6 3.06



which occurred after the actuation of C3 had ceased. Althoughmost prior work generating quantitative data has focused on theinitial exposure period and the return of the spine to the neutralposture [25], it is possible that the kinematics immediately follow-ing a dynamic exposure may actually be injurious if the magni-tude of deceleration is sufficient. Further, the facet compressionand separation in our studies were within 0.61 mm of the neutralposition in all the tests. The peak facet compression during rear-end collision simulation of �0.22 6 0.21 mm was lower than the�0.9 to 2.8 mm previously reported for 3.5–4.4 g impacts[25,28,38]. However, despite being small, the compressions in thecurrent study were significantly greater than during the quasi-static tests at an equivalent intervertebral rotation, which has notbeen previously reported. This suggests that while the present

testing may underestimate the absolute compression, an increasedrisk of an injury may exist.

Although there was little variation across the age of the speci-mens (mean of 66 years), it is possible that younger specimenswould exhibit different mechanical and physiological responsesunder the loading conditions of the present study. However, thespecimens tested represent a realistic population that may be sub-jected to such loading and are of a similar age to those tested inthe majority of in vitro cadaveric studies.

The dynamic loading exposures did not alter the overall kine-matics during quasi-static loading (Fig. 2). Rear-end impact simu-lations using both incremental accelerations of 2, 3.5, 5, 6.5, and8 g and a single 8 g exposure increase both the ROM and the neu-tral zone, suggesting that soft tissue injury may be induced [56].

Fig. 6 Representative full-field strain for specimen 1 ((a) and (b)) and specimen 4 ((c) and (d)).The peak MPS during dynamic flexion ((a) and (c)) and the MPS at equivalent C2/C3 rotationduring quasi-static testing ((b) and (d)) are shown. The variable MPS across the facet capsulewas determined; the arrows show the MPS direction within each element.

Table 3 Mean 6 SD maximum facet joint sliding, separation, compression, and linear strain

Test Parameter n 100 deg/s Quasi-static Significance

Flexion Posterior–anterior sliding (mm) 5 �1.01 6 0.46 �0.35 6 0.31 0.043a

Separation (positive) (mm) 6 0.20 6 0.18 0.18 6 0.19 0.600Compression (negative) (mm) 6 �0.22 6 0.21 �0.08 6 0.17 0.027a

Maximum strain across the facet (%) 5 3.92 6 1.48 3.38 6 2.56 0.345Extension Posterior–anterior sliding (mm) 5 1.19 6 0.46 0.44 6 0.56 0.042a

Separation (positive) (mm) 5 0.22 6 0.09 0.16 6 0.19 0.416Compression (negative) (mm) 5 �0.24 6 0.11 �0.18 6 0.18 0.343Maximum strain across the facet (%) 4 2.78 6 1.18 2.56 6 2.30 1.000

aA significant difference between dynamic and quasi-static tests (p< 0.05).



However, while the current study only assessed the rotation ateach vertebral level with respect to the global ROM in quasi-statictests that were performed before and after dynamic loading, thefact that there were no significant differences in the vertebral rota-tions indicates that soft tissue damage is not likely to haveoccurred in this study. Moreover, while it is possible that such anapproach would not identify soft tissue damage as accurately aswould be evident by a change in neutral zone, it is also possiblethat the relatively low accelerations and decelerations of 1.8–2.8 gused in dynamic exposures of the current study would not besufficient to elicit such a change.

The inertia of the specimen when the actuation ceased led to abrief period when the specimens adopted an S-shape (Fig. 7). Thisinertial effect in dynamic flexion corresponded approximately tothe point of the peak MPS, maximum SS, and anterior facet slid-ing (Fig. 4), all of which were significantly higher compared toquasi-static loading (Fig. 5; Tables 2 and 3). This inertial effect onthe upper cervical spine indicates that the deceleration followingan impact is critical in terms of facet mechanics relating to thepeak strain, and that minimizing the deceleration following a rear-end collision would limit the combination of C2/C3 rotation andanterior sliding that may lead to injury and pain. Previous studieshave reported that the induction of an S-shape to the cervical spineand altered vertebral centers of rotation and facet kinematics arelikely to be critical in the development of injury [25,28,36]. Thepresent results are consistent with this hypothesis, which demon-strate that the peak facet sliding corresponds with the peak MPSand SS. The altered kinematics combined with increased facetcapsule strain observed in the dynamic flexion tests compared toquasi-static loading suggest that loading to C2/C3 is different inthose two scenarios. The present study compares peak dynamicstrains with quasi-static strains at equivalent C2/C3 rotations;findings suggest that the significant differences observed here maybe related to altered loading which causes increased facet slidingand increases the likelihood of facet capsular ligament injury.However, other loading scenarios and their effects on the spinal

kinematics and full-field facet capsule strains should be investi-gated especially compared to physiological loading in order tofully understand how injuries may occur in difference collisionscenarios.

Although showing similar kinematic patterns relating to theinertial effect of the surrogate head as in dynamic flexion,dynamic extension simulating a frontal collision did not produceincreased MPS or SS following actuation (Fig. 4; Table 2). In con-trast, the maximum shear strain, which occurred during actuation(20–81 ms), was significantly higher than during quasi-static load-ing (Table 2). This may be due to the center of mass of the headnaturally falling into flexion, and thus limiting the extent to whichthe dynamic loading increased the MPS and SS. It is possible thatby implementing a follower-load to simulate passive muscleforces and to maintain the neutral position of the head, the MPSand SS would increase once the actuation stopped, similar to whatwas observed in the rear-end collision simulations of the presentstudy (Fig. 4).

The full-field strains showed that both MPS and SS were inho-mogeneous (Fig. 6), and that the full-field strain measurement ofthe capsule more accurately identifies local responses within thefacet capsule compared to measuring the linear strain across theentire facet joint. The adoption of the previously publishedmethod to assess overall facet behavior [25] did not reveal the sig-nificance in terms of increased strain during dynamic flexion(Table 3) that was identified using the full-field strain method(Fig. 5). These results are consistent with the previous study [25]but emphasize that measuring the linear strain across the entirefacet may underestimate the strain magnitude. The peak MPS dur-ing dynamic flexion (12.52 6 4.59%) approached the peak MPSthat has been reported in in vivo studies of facet capsule stretch ina rat model of facet-mediated painful mechanical injury [35].Interestingly, the linear strains measured across the entire facetjoint in the current study during dynamic flexion (Table 3) aresimilar to the peak MPS induced in a nonpainful in vivo group ofDong et al. (Table 5). The peak MPS measured during dynamic

Fig. 7 Mean rotation angle (flexion positive) during dynamic actuation applied at100 deg/s to the C3 level from 20 to 81 ms in either flexion (a) or extension (b). The vis-ual representations above the plots show the shape of the spine at 0, 40, 81, 120, and150 ms.

Table 4 Mean 6 standard deviation (SD) maximum and minimum rotations (flexion positive) at each cervical spinal level duringdynamic flexion for the current study in comparison to published cadaveric simulations of rear-end collisions

Present study (2.8 g) Ivancic et al. (3.5 g) [39] Grauer et al. (2.5 g) [54]

Level Max Min Max Min Max Min

Occ/C1 5.31 6 2.76 �0.39 6 0.37 4.9 6 3.7 �0.8 6 1.1 9.2 6 5.8 �13.4 6 7.5C1/C2 7.88 6 3.53 �0.02 6 0.03 2.0 6 1.6 �1.8 6 2.1 16.6 6 8.7 �13.8 6 3.6C2/C3 1.74 6 1.14 �1.54 6 0.98 1.5 6 1.1 �4.9 6 3.9 1.7 6 1.1 �7.0 6 2.6



flexion in the present study is similar to the strain in the C2/C3facet capsule during whiplash exposures to human cadaveric cer-vical spines [25,31] (Table 5). However, since the peak MPSmeasured from full-field strain analysis was much higher than thelinear strain across the facet during both dynamic flexion andextension in the present study (Table 3), it is likely that the lattermethod will fail to identify the peak strain, and may, therefore,underestimate the likelihood of injury, pain, and/or otherpathophysiological dysfunctional responses due to a traumaticexposure.

Pain can be induced at facet capsule strains much lower thanthe subcatastrophic failure strains that have been reported duringwhiplash at approximately 35–67% [32,40,41]. The notion that athreshold for pain production is lower than failure may be due tomicroscopic injury of the collagen fiber matrix [42,43,57,58]. Theonset of injury at the microstructural level has been identifiedusing quantitative polarized light imaging [42,43,57] and usingproxies such as afferents in the ligament [59–61]. The full-fieldstrain analysis of the present study does show that the peak MPSfrom dynamic flexion generally occurs in the anterior aspect ofthe facet capsule and is aligned along the facet joint (Fig. 6),which is consistent with increased anterior sliding at the time ofpeak MPS. Little correlation was found between fiber realignmentand the maximum MPS or SS in isolated capsules loaded in dis-traction [42,57], or posterior retraction [43]. It is possible thathigher resolution full-field strain analysis would better predict themagnitude and location of capsule injury, but this method wouldnot account for subsurface injury of the capsule that identifiedcapsule damage through collagen fiber realignment using QPLI.

The full-field shear strain in the cervical facet capsules (Fig. 4)has not previously been investigated in dynamic loading expo-sures representative of either frontal or rear-end collisions in mul-tilevel cervical specimens; yet, it has been measured during theretraction [43,62] and distraction [42] of isolated cervical facetjoints, and in functional spinal units undergoing pure moment test-ing [32]. There is limited shear strain data for the facet capsule atsubcatastrophic or catastrophic failure, making it difficult toassess the maximum and minimum shear strain of the presentstudy (Table 2) in relation to either injury or pain. The maximumand minimum SS in the present study were lower than measuredduring quasi-static loading [42,43]; but this may be due to boththe direction and rate of loading, which affect the location, magni-tude, and orientation of both MPS and SS. The adoption of meth-ods to measure local ligament kinematics and kinetics along withthe facet joint kinematics measured in the present study wouldprovide a greater understanding of how collagen fiber responsesand strains relate to capsule injury under loading conditions repre-sentative of traffic collisions and would assist in understandingwhich conditions increase the likelihood of injury.

Full-field strain measurements provide a quantitative measureof the inhomogeneous strain behavior of the facet capsule, whichallows a more accurate understanding of the risk of injury duringdynamic loading compared to linear strain analysis across theentire facet. The present study demonstrates that loading

representative of low-speed rear-end collisions produces injuriousMPS in the C2/C3 facet capsule, and that such exposures alter theanterior–posterior sliding motions of the facet joint and increasethe facet capsular MPS and SS relative to those during physiologicflexion–extension. Although upper cervical spine loading repre-sentative of low-speed frontal collisions leads to similar changesin anterior–posterior joint sliding and SS in the facet capsule com-pared to quasi-static flexion–extension, those loading conditionsdid not lead to increased MPS in the C2/C3 facet capsule.

Currently, there are no published reports on whether there arelevel-by-level differences in the anatomy and neurophysiology ofthe cervical spine that would lead to spinal level-specific facetcapsule injury thresholds under dynamic loading conditions.Additional in vivo studies could provide deeper understanding ofwhether there are potential differences in strain and/or painthresholds for the facet capsule or different levels of the cervicalspine. Furthermore, it remains difficult to identify the microstruc-tural injury mechanisms within the facet during dynamic loadingconditions representative of in vivo scenarios, such as traffic colli-sions. But the significantly elevated strains under dynamic expo-sures compared to quasi-static loading in the present studysuggest that strains in the C2/C3 facet capsule may reach levelssufficient to induce pain and/or other local injury. Future workcombining these data with microstructural analyses of isolatedfacet capsules and in vivo studies could help fully define howsuch loading exposures occur, and what loading conditions put thecervical spine at the greatest risk of injury. The results also dem-onstrate that the deceleration following a rear-end collision is justas critical in avoiding injury as the collision itself, and the abilityto minimize such decelerations through improved safety mecha-nisms may reduce the likelihood of neck pain following suchexposures.

Acknowledgment

The authors gratefully acknowledge the Catherine SharpeFoundation for providing funding support to this research. TheCatherine Sharpe Foundation provided financial support but wasnot involved in the design, completion, analysis, or preparationand submission of this article.

References[1] Siegmund, G. P., Winkelstein, B. A., Ivancic, P. C., Svensson, M. Y., and Vasa-

vada, A., 2009, “The Anatomy and Biomechanics of Acute and Chronic Whip-lash Injury,” Traffic Inj. Prev., 10(2), pp. 101–112.

[2] Jaumard, N. V., Welch, W. C., and Winkelstein, B. A., 2011, “Spinal FacetJoint Biomechanics and Mechanotransduction in Normal, Injury and Degenera-tive Conditions,” ASME J. Biomech. Eng., 133(7), p. 071010.

[3] Lord, S. M., Barnsley, L., Wallis, B. J., and Bogduk, N., 1996, “Chronic Cervi-cal Zygapophysial Joint Pain After Whiplash. A Placebo-Controlled PrevalenceStudy,” Spine, 21(15), pp. 1737–1744; discussion 1744–1735.

[4] Quinlan, K. P., Annest, J. L., Myers, B., Ryan, G., and Hill, H., 2004, “NeckStrains and Sprains Among Motor Vehicle Occupants—United States, 2000,”Accid. Anal. Prev., 36(1), pp. 21–27.

Table 5 Summary of mean 6 SD facet strain data from painful and nonpainful in vivo studies and during quasi-static (QS) anddynamic In vitro loading

Study Measurement Test method Level Variable Strain (%)

Dong et al. [35] 2D full-field strain (MPS) In vivo rat facet retraction C6/C7 Painful 19.43 6 11.43Nonpainful 6.29 6 2.64

Present study 3D full-field strain (MPS) In vitro rear-end collision C2/C3 2.8 g 12.52 6 4.59QS 7.63 6 4.35

Linear strain of facet In vitro rear-end collision C2/C3 2.8 g 3.92 6 1.33QS 3.38 6 2.56

Pearson et al. [25] Linear strain of facet In vitro whiplash C2/C3 3.5 g 13.4 6 9.3QS 9.4 6 7.1

Panjabi et al. [31] Linear strain of facet In vitro whiplash C2/C3 2.5 g 15.9 6 20.0



http://dx.doi.org/10.1080/15389580802593269

http://dx.doi.org/10.1115/1.4004493

http://dx.doi.org/10.1097/00007632-199608010-00005

http://dx.doi.org/10.1016/S0001-4575(02)00110-0

[5] Zuby, D. S., and Lund, A. K., 2010, “Preventing Minor Neck Injuries in RearCrashes—Forty Years of Progress,” J. Occup. Environ. Med., 52(4),pp. 428–433.

[6] White, K., Hudgins, T. H., and Alleva, J. T., 2009, “Cervical Facet MediatedPain,” Disease, 55(12), pp. 729–736.

[7] Schofferman, J., Bogduk, N., and Slosar, P., 2007, “Chronic Whiplash andWhiplash-Associated Disorders: An Evidence-Based Approach,” J. Am. Acad.Orthop. Surg., 15(10), pp. 596–606.

[8] Radanov, B. P., Sturzenegger, M., and Di Stefano, G., 1995, “Long-Term Out-come After Whiplash Injury. A 2-Year Follow-Up Considering Features ofInjury Mechanism and Somatic, Radiologic, and Psychosocial Findings,” Medi-cine, 74(5), pp. 281–297.

[9] Gellhorn, A. C., 2011, “Cervical Facet-Mediated Pain,” Phys. Med. Rehab.Clin. N. Am., 22(3), pp. 447–458.

[10] Manchikanti, L., Singh, V., Rivera, J., and Pampati, V., 2002, “Prevalence ofCervical Facet Joint Pain in Chronic Neck Pain,” Pain Physician, 5(3), pp.243–249.

[11] Berglund, A., Alfredsson, L., Cassidy, J. D., Jensen, I., and Nygren, A., 2000,“The Association Between Exposure to a Rear-End Collision and Future Neckor Shoulder Pain: A Cohort Study,” J. Clin. Epidemiol., 53(11), pp. 1089–1094.

[12] Bogduk, N., and Marsland, A., 1988, “The Cervical Zygapophysial Joints as aSource of Neck Pain,” Spine, 13(6), pp. 610–617.

[13] Fukui, S., Ohseto, K., Shiotani, M., Ohno, K., Karasawa, H., Naganuma, Y.,and Yuda, Y., 1996, “Referred Pain Distribution of the Cervical ZygapophysealJoints and Cervical Dorsal Rami,” Pain, 68(1), pp. 79–83.

[14] Barnsley, L., Lord, S. M., Wallis, B. J., and Bogduk, N., 1995, “The Prevalenceof Chronic Cervical Zygapophysial Joint Pain After Whiplash,” Spine, 20(1),pp. 20–25; discussion 26.

[15] Chen, H. B., Yang, K. H., and Wang, Z. G., 2009, “Biomechanics of WhiplashInjury,” Chin. J. Traumatol., 12(5), pp. 305–314.

[16] Winkelstein, B. A., 2011, “How Can Animal Models Inform on the Transitionto Chronic Symptoms in Whiplash?,” Spine, 36(Suppl. 25), pp. S218–S225.

[17] Chen, C., Lu, Y., Cavanaugh, J. M., Kallakuri, S., and Patwardhan, A., 2005,“Recording of Neural Activity From Goat Cervical Facet Joint Capsule UsingCustom-Designed Miniature Electrodes,” Spine, 30(12), pp. 1367–1372.

[18] Inami, S., Shiga, T., Tsujino, A., Yabuki, T., Okado, N., and Ochiai, N., 2001,“Immunohistochemical Demonstration of Nerve Fibers in the Synovial Fold ofthe Human Cervical Facet Joint,” J. Orthop. Res., 19(4), pp. 593–596.

[19] McLain, R. F., 1994, “Mechanoreceptor Endings in Human Cervical FacetJoints,” Spine, 19(5), pp. 495–501.

[20] Kras, J. V., Tanaka, K., Gilliland, T. M., and Winkelstein, B. A., 2013, “AnAnatomical and Immunohistochemical Characterization of Afferents Innervat-ing the C6-C7 Facet Joint After Painful Joint Loading in the Rat,” Spine, 38(6),pp. E325–E331.

[21] Kras, J. V., Weisshaar, C. L., Pall, P. S., and Winkelstein, B. A., 2015, “PainFrom Intra-Articular NGF or Joint Injury in the Rat Requires ContributionsFrom Peptidergic Joint Afferents,” Neurosci. Lett., 604, pp. 193–198.

[22] Cavanaugh, J. M., Lu, Y., Chen, C., and Kallakuri, S., 2006, “Pain Generationin Lumbar and Cervical Facet Joints,” J. Bone Joint Surg. Am., 88(Suppl. 2),pp. 63–67.

[23] Kallakuri, S., Singh, A., Lu, Y., Chen, C., Patwardhan, A., and Cavanaugh, J.M., 2008, “Tensile Stretching of Cervical Facet Joint Capsule and Related Axo-nal Changes,” Eur. Spine J., 17(4), pp. 556–563.

[24] Crosby, N. D., Gilliland, T. M., and Winkelstein, B. A., 2014, “Early AfferentActivity From the Facet Joint After Painful Trauma to Its Capsule PotentiatesNeuronal Excitability and Glutamate Signaling in the Spinal Cord,” Pain,155(9), pp. 1878–1887.

[25] Pearson, A. M., Ivancic, P. C., Ito, S., and Panjabi, M. M., 2004, “Facet JointKinematics and Injury Mechanisms During Simulated Whiplash,” Spine, 29(4),pp. 390–397.

[26] Tominaga, Y., Ndu, A. B., Coe, M. P., Valenson, A. J., Ivancic, P. C., Ito, S.,Rubin, W., and Panjabi, M. M., 2006, “Neck Ligament Strength is DecreasedFollowing Whiplash Trauma,” BMC Musculoskeletal Disord., 7, p. 103.

[27] Winkelstein, B. A., McLendon, R. E., Barbir, A., and Myers, B. S., 2001, “AnAnatomical Investigation of the Human Cervical Facet Capsule, QuantifyingMuscle Insertion Area,” J. Anat., 198(Pt. 4), pp. 455–461.

[28] Cusick, J. F., Pintar, F. A., and Yoganandan, N., 2001, “Whiplash Syndrome:Kinematic Factors Influencing Pain Patterns,” Spine, 26(11), pp. 1252–1258.

[29] Ivancic, P. C., Panjabi, M. M., Ito, S., Cripton, P. A., and Wang, J. L., 2005,“Biofidelic Whole Cervical Spine Model With Muscle Force Replication forWhiplash Simulation,” Eur. Spine J., 14(4), pp. 346–355.

[30] Dehner, C., Elbel, M., Schick, S., Walz, F., Hell, W., and Kramer, M., 2007,“Risk of Injury of the Cervical Spine in Sled Tests in Female Volunteers,” Clin.Biomech., 22(6), pp. 615–622.

[31] Panjabi, M. M., Cholewicki, J., Nibu, K., Grauer, J., and Vahldiek, M., 1998,“Capsular Ligament Stretches During In Vitro Whiplash Simulations,” J. SpinalDisord., 11(3), pp. 227–232.

[32] Winkelstein, B. A., Nightingale, R. W., Richardson, W. J., and Myers, B. S.,2000, “The Cervical Facet Capsule and Its Role in Whiplash Injury: A Biome-chanical Investigation,” Spine, 25(10), pp. 1238–1246.

[33] Anderst, W. J., Donaldson, W. F., 3rd, Lee, J. Y., and Kang, J. D., 2014,“In Vivo Cervical Facet Joint Capsule Deformation DuringFlexion–Extension,” Spine, 39(8), pp. E514–E520.

[34] Kras, J. V., Dong, L., and Winkelstein, B. A., 2014, “Increased Interleukin-1alpha and Prostaglandin E2 Expression in the Spinal Cord at 1 Day After Pain-ful Facet Joint Injury: Evidence of Early Spinal Inflammation,” Spine, 39(3),pp. 207–212.

[35] Dong, L., Quindlen, J. C., Lipschutz, D. E., and Winkelstein, B. A., 2012,“Whiplash-Like Facet Joint Loading Initiates Glutamatergic Responses in theDRG and Spinal Cord Associated With Behavioral Hypersensitivity,” BrainRes., 1461, pp. 51–63.

[36] Bogduk, N., and Yoganandan, N., 2001, “Biomechanics of the CervicalSpine—Part 3: Minor Injuries,” Clin. Biomech., 16(4), pp. 267–275.

[37] Luan, F., Yang, K. H., Deng, B., Begeman, P. C., Tashman, S., and King, A. I.,2000, “Qualitative Analysis of Neck Kinematics During Low-Speed Rear-EndImpact,” Clin. Biomech., 15(9), pp. 649–657.

[38] Yoganandan, N., Pintar, F. A., and Cusick, J. F., 2002, “Biomechanical Analy-ses of Whiplash Injuries Using an Experimental Model,” Accid. Anal. Prev.,34(5), pp. 663–671.

[39] Ivancic, P. C., Panjabi, M. M., and Ito, S., 2006, “Cervical Spine Loads andIntervertebral Motions During Whiplash,” Traffic Inj. Prev., 7(4), pp. 389–399.

[40] Siegmund, G. P., Myers, B. S., Davis, M. B., Bohnet, H. F., and Winkelstein, B.A., 2001, “Mechanical Evidence of Cervical Facet Capsule Injury DuringWhiplash: A Cadaveric Study Using Combined Shear, Compression, andExtension Loading,” Spine, 26(19), pp. 2095–2101.

[41] Winkelstein, B. A., Nightingale, R. W., Richardson, W. J., and Myers, B. S.,1999, “Cervical Facet Joint Mechanics: Its Application to Whiplash Injury,”Stapp Car Crash J., 43, pp. 243–252.

[42] Quinn, K. P., and Winkelstein, B. A., 2008, “Altered Collagen Fiber KinematicsDefine the Onset of Localized Ligament Damage During Loading,” J. Appl.Physiol., 105(6), pp. 1881–1888.

[43] Quinn, K. P., and Winkelstein, B. A., 2011, “Detection of Altered CollagenFiber Alignment in the Cervical Facet Capsule After Whiplash-Like JointRetraction,” Ann. Biomed. Eng., 39(8), pp. 2163–2173.

[44] Lu, Y., Chen, C., Kallakuri, S., Patwardhan, A., and Cavanaugh, J. M., 2005,“Neural Response of Cervical Facet Joint Capsule to Stretch: A Study of Whip-lash Pain Mechanism,” Stapp Car Crash J., 49, pp. 49–65.

[45] Yoganandan, N., Pintar, F. A., and Klienberger, M., 1998, “Cervical Spine Ver-tebral and Facet Joint Kinematics Under Whiplash,” ASME J. Biomech. Eng.,120(2), pp. 305–307.

[46] Panjabi, M. M., Cholewicki, J., Nibu, K., Babat, L. B., and Dvorak, J., 1998,“Simulation of Whiplash Trauma Using Whole Cervical Spine Specimens,”Spine, 23(1), pp. 17–24.

[47] Descarreaux, M., Blouin, J. S., and Teasdale, N., 2003, “A Non-Invasive Tech-nique for Measurement of Cervical Vertebral Angle: Report of a PreliminaryStudy,” Eur. Spine J., 12(3), pp. 314–319.

[48] Yoganandan, N., Pintar, F. A., Zhang, J., and Baisden, J. L., 2009, “PhysicalProperties of the Human Head: Mass, Center of Gravity and Moment ofInertia,” J. Biomech., 42(9), pp. 1177–1192.

[49] Deng, B., Begeman, P. C., Yang, K. H., Tashman, S., and King, A. I., 2000,“Kinematics of Human Cadaver Cervical Spine During Low Speed Rear-EndImpacts,” Stapp Car Crash J., 44, pp. 171–188.

[50] Bogduk, N., and Mercer, S., 2000, “Biomechanics of the Cervical Spine—I:Normal Kinematics,” Clin. Biomech., 15(9), pp. 633–648.

[51] Anderst, W. J., 2015, “Bootstrap Prediction Bands for Cervical Spine Interver-tebral Kinematics During In Vivo Three-Dimensional Head Movements,”J. Biomech., 48(7), pp. 1270–1276.

[52] Bostrom, O., Fredriksson, R., Haland, Y., Jakobsson, L., Krafft, M., Lovsund,P., Muser, M. H., and Svensson, M. Y., 2000, “Comparison of Car Seats in LowSpeed Rear-End Impacts Using the BioRID Dummy and the New Neck InjuryCriterion (NIC),” Accid. Anal. Prev., 32(2), pp. 321–328.

[53] Dehner, C., Schick, S., Kraus, M., Hell, W., and Kramer, M., 2013, “MuscleActivity Influence on the Kinematics of the Cervical Spine in Rear-End SledTests in Female Volunteers,” Traffic Inj. Prev., 14(4), pp. 369–377.

[54] Grauer, J. N., Panjabi, M. M., Cholewicki, J., Nibu, K., and Dvorak, J., 1997,“Whiplash Produces an S-Shaped Curvature of the Neck With Hyperextensionat Lower Levels,” Spine, 22(21), pp. 2489–2494.

[55] Stemper, B. D., Yoganandan, N., Gennarelli, T. A., and Pintar, F. A., 2005,“Localized Cervical Facet Joint Kinematics Under Physiological and WhiplashLoading,” J. Neurosurg, 3(6), pp. 471–476.

[56] Ghole, S. A., Ivancic, P. C., Tominaga, Y., Gimenez, S. E., and Panjabi, M. M.,2004, “Incremental and Single Trauma Produce Equivalent Subfailure Soft Tis-sue Injury of the Cervical Spine,” Clin. Biomech., 19(8), pp. 784–789.

[57] Quinn, K. P., and Winkelstein, B. A., 2009, “Vector Correlation Technique forPixel-Wise Detection of Collagen Fiber Realignment During Injurious TensileLoading,” J. Biomed. Opt., 14(5), p. 054010.

[58] Zhang, S., Cao, X., Stablow, A. M., Shenoy, V. B., and Winkelstein, B. A., 2016,“Tissue Strain Reorganizes Collagen With a Switch-Like Response That Regu-lates Neuronal ERK Phosphorylation In Vitro: Implications for LigamentousInjury and Mechanotransduction,” ASME J. Biomech. Eng., 138(2), p. 021013.

[59] Lu, Y., Chen, C., Kallakuri, S., Patwardhan, A., and Cavanaugh, J. M., 2005,“Neurophysiological and Biomechanical Characterization of Goat CervicalFacet Joint Capsules,” J. Orthop. Res., 23(4), pp. 779–787.

[60] Lu, Y., Chen, C., Kallakuri, S., Patwardhan, A., and Cavanaugh, J. M., 2005,“Development of an In Vivo Method to Investigate Biomechanical and Neuro-physiological Properties of Spine Facet Joint Capsules,” Eur. Spine J., 14(6),pp. 565–572.

[61] Avramov, A. I., Cavanaugh, J. M., Ozaktay, C. A., Getchell, T. V., and King,A. I., 1992, “The Effects of Controlled Mechanical Loading on Group-II, III,and IV Afferent Units From the Lumbar Facet Joint and Surrounding Tissue.An In Vitro Study,” J. Bone Joint Surg., 74(10), pp. 1464–1471.

[62] Lee, D. J., and Winkelstein, B. A., 2012, “The Failure Response of the HumanCervical Facet Capsular Ligament During Facet Joint Retraction,” J. Biomech.,45(14), pp. 2325–2329.



http://dx.doi.org/10.1097/JOM.0b013e3181bb777c

http://dx.doi.org/10.1016/j.disamonth.2009.06.003

http://dx.doi.org/10.5435/00124635-200710000-00004

http://dx.doi.org/10.5435/00124635-200710000-00004

http://dx.doi.org/10.1097/00005792-199509000-00005

http://dx.doi.org/10.1097/00005792-199509000-00005

http://dx.doi.org/10.1016/j.pmr.2011.02.006

http://dx.doi.org/10.1016/j.pmr.2011.02.006

http://www.painphysicianjournal.com/linkout?issn=1533-3159&vol=5&page=243

http://dx.doi.org/10.1016/S0895-4356(00)00225-0

http://dx.doi.org/10.1097/00007632-198813060-00003

http://dx.doi.org/10.1016/S0304-3959(96)03173-9

http://dx.doi.org/10.1097/00007632-199501000-00004

http://dx.doi.org/10.3760/cma.j.issn.1008-1275.2009.05.011

http://dx.doi.org/10.1097/BRS.0b013e3182387f96

http://dx.doi.org/10.1097/01.brs.0000166193.39389.21

http://dx.doi.org/10.1016/S0736-0266(00)00048-6

http://dx.doi.org/10.1097/00007632-199403000-00001

http://dx.doi.org/10.1097/BRS.0b013e318285b5bb

http://dx.doi.org/10.1016/j.neulet.2015.07.043

http://dx.doi.org/10.2106/JBJS.E.01411

http://dx.doi.org/10.1007/s00586-007-0562-0

http://dx.doi.org/10.1016/j.pain.2014.06.019

http://dx.doi.org/10.1097/01.BRS.0000090836.50508.F7

http://dx.doi.org/10.1186/1471-2474-7-103

http://dx.doi.org/10.1017/S0021878201007518

http://dx.doi.org/10.1097/00007632-200106010-00015

http://dx.doi.org/10.1007/s00586-004-0758-5

http://dx.doi.org/10.1016/j.clinbiomech.2007.02.004


http://dx.doi.org/10.1097/00007632-200005150-00007

http://dx.doi.org/10.1097/BRS.0000000000000235

http://dx.doi.org/10.1097/BRS.0000000000000107

http://dx.doi.org/10.1016/j.brainres.2012.04.026

http://dx.doi.org/10.1016/j.brainres.2012.04.026

http://dx.doi.org/10.1016/S0268-0033(01)00003-1

http://dx.doi.org/10.1016/S0268-0033(00)00031-0

http://dx.doi.org/10.1016/S0001-4575(01)00066-5

http://dx.doi.org/10.1080/15389580600789127

http://dx.doi.org/10.1097/00007632-200110010-00010

http://papers.sae.org/99SC15/

http://dx.doi.org/10.1152/japplphysiol.90792.2008

http://dx.doi.org/10.1152/japplphysiol.90792.2008

http://dx.doi.org/10.1007/s10439-011-0316-3

http://papers.sae.org/2005-22-0003/

http://dx.doi.org/10.1115/1.2798318

http://dx.doi.org/10.1097/00007632-199801010-00005

http://dx.doi.org/10.1007/s00586-002-0511-x

http://dx.doi.org/10.1016/j.jbiomech.2009.03.029

http://papers.sae.org/2000-01-sc13/

http://dx.doi.org/10.1016/S0268-0033(00)00034-6


http://dx.doi.org/10.1016/S0001-4575(99)00105-0

http://dx.doi.org/10.1080/15389588.2012.714018

http://dx.doi.org/10.1097/00007632-199711010-00005

http://dx.doi.org/10.3171/spi.2005.3.6.0471


http://dx.doi.org/10.1117/1.3227037

http://dx.doi.org/10.1115/1.4031975

http://dx.doi.org/10.1016/j.orthres.2005.01.002

http://dx.doi.org/10.1007/s00586-004-0835-9

http://jbjs.org/content/74/10/1464.long


Upper Cervical Spine Loading Simulating a Dynamic Low-Speed …winkelst/Holsgrove2016.pdf · 2016....

Documents

Transcript of Upper Cervical Spine Loading Simulating a Dynamic Low-Speed …winkelst/Holsgrove2016.pdf · 2016....